Court records are used extensively in historical research. Preserved as summaries of daily legal proceedings, they give historians a unique opportunity to access information about the names, characteristics, and socio‐economic status of individuals and the laws, local customs, and legal institutions of societies. Although researchers have noted various limitations of these records, the problem of selection bias has not been systematically studied. Since litigants would probably settle disputes in which one side is likely to be a clear winner, the cases that go to trial are more likely to be the difficult and uncertain ones that comprise a non‐random subset of all disputes. This article presents a study of selection bias in Ottoman courts in the town of Kastamonu in northern Anatolia, from the late seventeenth and the eighteenth centuries. Disputes are categorized by type and the distribution of court participants is studied according to composition, gender, and socio‐religious status. A regression analysis is run to determine the factors affecting the likelihood of cases being tried in court. The results indicate that the cases that ended up in court were selected systematically. If the selection bias is ignored, research based on Ottoman court records may be seriously flawed in its ability to yield general conclusions.